Bimetallic Synergistic Catalysis Enabling Efficient Chemical Recycling of Commercial Polyesters

Abstract

Poly­(ethylene terephthalate) (PET) is a globally mass-produced polyester that is vital to many industries, but its accumulation poses serious environmental challenges, necessitating efficient chemical recycling. However, the efficiency of these recycling strategies is limited by their inadequate catalytic performance. Herein, we developed a highly effective mononuclear Zn catalyst that achieves PET glycolysis with an activity of 1800 g_PET g_cat ^–1 h^–1. Mechanistic investigations revealed that this complex serves as a precatalyst and undergoes in situ dimerization to generate dinuclear Zn species as the true active species. Most importantly, polyester depolymerization proceeds via a bimetallic synergetic catalytic mechanism, wherein the two Zn centers within a unified framework synergistically activate the carbonyl substrate and alkoxide species, thereby significantly facilitating nucleophilic addition. This catalytic system exhibits broad applicability, achieving >95% depolymerization efficiency for various polyesters and postconsumer plastic blends.

Introduction

Poly­(ethylene terephthalate) (PET) is a widely employed thermoplastic material with various applications, including textile fibers and food packaging, owing to its favorable mechanical strength, chemical resistance, and light weight. , However, this widespread convenience leads to significant consumption of these polymers, resulting in the continuous accumulation of waste in ecosystems because of inadequate waste disposal. Most of this waste ultimately ends up in landfills or is incinerated, posing significant threats to ecosystems and potentially causing long-term effects on human health. Therefore, substantial efforts have been made to develop strategies for recycling PET waste. − Despite these advancements, the economic feasibility of recycling technologies remains challenging, primarily due to high energy cost and low efficiency. ,

Mechanical processing is widely used for plastic recycling; however, it faces several challenges, including material downgrading owing to shortened polymer chains and sensitivity to contamination, which severely limits its application in high-value products. , Alternatively, the chemical recycling of PET waste to its constituent monomers has emerged as a pivotal approach to sustainable polymer management. This strategy not only diminishes the consumption of petroleum-based feedstocks but also offers a permanent solution to plastic waste. Several chemical recycling approaches including hydrolysis, methanolysis, glycolysis, and hydrogenation have been developed for PET. − Among them, PET glycolysis to bis­(2-hydroxyethyl) terephthalate (BHET) has gained prominence as a particularly promising route, generally operating under mild conditions with low-volatility diols at ambient pressure. Nevertheless, the effectiveness of this process depends critically on the development of high-performance catalysts that are crucial for enabling efficient and economically viable PET recycling. Consequently, the development of highly efficient catalysts is crucial for achieving sustainable PET recycling.

The depolymerization of acyl-based polymers involves the nucleophilic addition of alkoxide species to the carbonyl group and a subsequent elimination process. Therefore, the introduction of Lewis acids as catalysts is viable for facilitating acyl polymer depolymerization. , Consequently, remarkable efforts have been made for developing catalysts for the efficient depolymerization of acyl-based polymers. Williams et al. chemically recycled poly­(cyclohexene carbonate) into epoxides with CO₂ using a heterometallic Mg­(II)­Co­(II) complex. In particular, a series of dinuclear and multinuclear catalysts have been designed to enhance the efficiency of PET depolymerization. Representative systems, including biomimetic dinuclear Mn and multinuclear Ti phenoxide catalysts, have been developed to facilitate the chemical recycling of PET. (Figure a,b, respectively). , Furthermore, Dove et al. recently demonstrated a highly stable heterodinuclear M₁(II)­M₂(I) catalyst for efficient PET glycolysis (Figure c). Despite these advances, many existing catalytic systems require sophisticated pre-synthesis of well-defined structures and are expensive.

1.

(a) Biomimetic dinuclear Mn catalyst for PET glycolysis. (b) Multinuclear Ti phenoxide catalyst for PET alcoholysis. (c) Heterodinuclear M₁(II)­M₂(I) catalyst for PET glycolysis. (d) Mononuclear Zn catalyst generates in situ dimeric species that mediate PET glycolysis through a bimetallic synergistic catalysis (this work).

Generally, in the Lewis-acid-catalyzed depolymerization of PET, the rate-determining step is the nucleophilic addition of an alkoxide to the carbonyl group. The efficiency of this step is governed by two key factors: (i) dissociation of the alkoxide nucleophile from the metal center and (ii) its subsequent attack on the electrophilic carbonyl group of the ester linkage. We hypothesized that bimetallic synergistic catalysis can integrate the activation of the carbonyl group and the binding of alkoxide species into a limited domain, thereby enabling efficient PET glycolysis (Figure d). In this study, we developed a mononuclear Zn catalyst that generates dimeric Zn species in situ during PET depolymerization. This catalytic system enables efficient PET glycolysis via a synergistic dual-activation mechanism, leading to high activity and BHET selectivity. Notably, the catalytic system demonstrates broad substrate compatibility and excellent stability, underscoring its potential as a practical and universal strategy for advancing a circular polymer economy.

Results and Discussion

The mononuclear Zn catalyst 1 was synthesized for bulk PET glycolysis (Figure a). We initially investigated the effect of reaction temperature on the activity of 1-catalyzed PET glycolysis (Figure b). Mononuclear Zn catalyst 1 is capable of efficiently mediating PET glycolysis with a depolymerization activity of 467 g_PET g_cat ^–1 h^–1, while maintaining a selectivity of >99% for BHET; this was achieved with a PET/1 feed ratio of 1000/1 at 180 °C (Table S1, Entry 1). Lowering the temperature to 170 °C and 160 °C resulted in approximately 3-fold (156 g_PET g_cat ^–1 h^–1) and 5.5-fold (84 g_PET g_cat ^–1 h^–1) decreases in activity, respectively (Table S1, Entries 2 and 3). Notably, despite the reduced activity, no byproduct formation was observed, and a selectivity of >99% for BHET was maintained. Increasing the depolymerization temperature to 190 °C significantly enhanced the activity to 874 g_PET g_cat ^–1 h^–1 (Table S1, Entry 4). Further elevation to 200 °C, slightly above the boiling temperature of ethylene glycol (EG), achieved complete PET conversion within 25 min (activity = 1153 g_PET g_cat ^–1 h^–1), while retaining a >99% BHET selectivity (Table S1, Entry 5). The effect of catalyst concentration was also investigated at 200 °C (Figure c). Increasing the feed ratio of [PET]₀/[1]₀ to 2000:1 enhanced the depolymerization activity to 1373 g_PET g_cat ^–1 h^–1 (Table S1, Entry 6). At an even higher ratio of 4000:1, the activity further increased to 1538 g_PET g_cat ^–1 h^–1 with high selectivity (Table S1, Entry 7). Remarkably, 1 maintained its high performance at a feed ratio of [PET]₀/[1]₀ =8000/1, delivering an activity of 1730 g_PET g_cat ^–1 h^–1 (Table S1, Entry 8). Critically, even at an exceptionally high feed ratio of [PET]₀/[1]₀ =10000/1, 1 demonstrated undiminished activity (1800 g_PET g_cat ^–1 h^–1) and exceptional BHET selectivity (>99%), unequivocally demonstrating its outstanding robustness under harsh conditions (Table S1, Entry 9). These results indicate that 1 exhibited exceptional activity with >99% selectivity, outperforming the previously reported state-of-the-art catalysts for PET glycolysis (Table S2, see the Supporting Information for details). Moreover, conventional Zn-based glycolysis catalysts, including zinc acetate and zinc ethanolate, exhibited drastically lower activities under identical conditions (Figure S1), demonstrating the exceptional performance of catalyst 1 at low loadings.

2.

Catalytic performance of mononuclear Zn catalyst 1 for catalyzing PET glycolysis. (a) Structure of the mononuclear Zn catalyst 1. (b) Activities and selectivities for 1-catalyzed PET glycolysis at different temperatures. (c) Activities and selectivities for 1-catalyzed PET glycolysis at various catalyst loadings and 200 °C. (d) Determination of the dependence of reaction rate on catalyst concentration and logarithmic plots of the initial rate versus catalyst concentration for 1-catalyzed methyl benzoate glycolysis.

To better understand the origin of the exceptional performance of 1-mediated PET glycolysis, we conducted kinetic studies by monitoring the 1-catalyzed glycolysis of methyl benzoate, a model terephthalate, at varying catalyst concentrations (Figure d). Kinetic analysis revealed a reaction order of 2.01 for 1-catalyzed glycolysis, demonstrating a second-order dependence on catalyst concentration. This suggests that 1-catalyzed glycolysis proceeds through an intermolecular bimetallic synergistic catalysis mechanism. Generally, such systems would render the catalytic activity and selectivity highly susceptible to dilution, causing a sharp decline in efficiency at low catalyst loadings. − Therefore, we hypothesized that the mononuclear structure might form dinuclear species in situ during the depolymerization process, which would enable intramolecular bimetallic synergistic catalysis. Correspondingly, this effectively circumvents the inherent limitations of intermolecular systems imposed by dilution effects at low catalyst loadings, thereby supporting the exceptional efficiency of catalytic PET glycolysis. To investigate the potential pathways for the in situ formation of the dinuclear species of 1, we performed density functional theory (DFT) calculations for two possible formation mechanisms: (i) formation of complex D1 via a 2:1 stoichiometric ratio of 1 to EG; (ii) formation of complex D2 via a 1:1 stoichiometric ratio of 1 to EG (Figure ). The geometry-optimized structures of the mononuclear Zn catalyst were used to construct the Gibbs free energy profiles for each complete reaction pathway, including key intermediates and transition states (Figure S2). Comparative analysis indicated that the dimeric structure formed via pathway (ii) is thermodynamically more favorable (ΔG = −21.1 kcal/mol) than that formed via pathway (i) (ΔG = −1.8 kcal/mol). Furthermore, pathway (ii) exhibits a significant 6.6 kcal/mol reduction in the barrier for the formation of a dinuclear species. These results demonstrate that pathway (ii) is both kinetically and thermodynamically favored for the formation of the active dinuclear species from pre-catalyst 1. This computational result is further corroborated by three experimental observations: (1) treatment of 1 with excess EG (6 equiv) led to the detection of ethanol, a byproduct arising from anion ligand exchange (Figures S3 and S4); (2) the dinuclear intermediate 1a was isolated and characterized, featuring two hydroxyethoxy anion ligands bridging the Zn centers and one methyl benzoate molecule coordinated to one metal center per dinuclear unit (Figure S5); (3) both isolated 1a and complex D2 exhibited catalytic activity and >99% selectivity comparable to those of 1 under identical conditions (Figures S6–S8). Collectively, these experimental findings, in conjunction with the computational data, provide compelling evidence that the dinuclear species are fully competent catalytic entities, strongly supporting their role as the true active species.

3.

Gibbs free energy profile (with intermediates and transition states) for the potential pathways for the in situ formation of a dinuclear species of 1 (Gibbs free energies are given in kcal/mol).

We further investigated the depolymerization mechanism of PET. Depolymerization occurs primarily through two mechanisms: chain-end unzipping and random chain scission. To determine the mechanism of 1-mediated PET glycolysis, we monitored PET glycolysis by periodic ¹H NMR spectroscopy and gel permeation chromatography. A sharp decrease in the molecular weight was observed during the initial stages of the reaction, followed by a gradual decline and a temporary broadening of the dispersity, which later narrowed (Figure S9). This behavior confirms that 1-mediated PET depolymerization occurs via a random chain scission mechanism rather than chain-end unzipping.

In terms of 1-catalyzed PET glycolysis, DFT simulations indicate that the monometallic catalysis mechanism is thermodynamically unfavorable (ΔG = 1.5 kcal/mol), in agreement with the kinetics results (Figure d and Figure S10). We further performed DFT calculations to elucidate the complete catalytic cycle of 1-catalyzed PET depolymerization using methyl benzoate as a model compound for PET (Figure a). The cycle is initiated with dimerization to form the dinuclear intermediate INT-6. Next, Zn–O bond dissociation enables methyl benzoate coordination via an endergonic step (ΔG = 12.0 kcal/mol), yielding INT-1, a structural analog of the experimentally characterized complex 1a. Subsequently, nucleophilic addition occurs via the rate-determining transition state TS-1 (ΔG ^‡ = 16.8 kcal/mol), forming INT-2 in an exergonic step (ΔG = −3.9 kcal/mol). Moreover, the rate-determining step of 1-catalyzed PET depolymerization via an intramolecular bimetallic synergistic catalysis pathway has an energy barrier 3.7 kcal/mol lower than that of the intermolecular pathway (Figure b and Figure S11). This significant energy difference is a key factor that contributes to the exceptional depolymerization efficiency of catalyst 1. Elimination via TS-2 (ΔG ^‡ = 4.6 kcal/mol) generates INT-3 (ΔG = −0.6 kcal/mol), followed by exergonic release of 2-hydroxyethyl benzoate and EG coordination to form INT-4 (ΔG = −8.5 kcal/mol). A subsequent low-barrier proton transfer (TS-3, ΔG ^‡ = 0.1 kcal/mol) affords INT-5 (ΔG = −0.4 kcal/mol). Cycle completion occurs with the endergonic release of methanol (ΔG = 1.4 kcal/mol), regenerating the dinuclear species INT-6 for subsequent catalytic depolymerization cycles.

4.

(a) Calculated catalytic cycle mechanism for the dimeric Zn species-mediated PET glycolysis via an intramolecular bimetallic synergistic catalysis mechanism. (b) The rate-determining step of 1-mediated PET glycolysis via an intermolecular bimetallic synergistic catalysis mechanism. (c) Charge analyses of the two bimetallic pathways. The Gibbs free energies are given in kcal/mol.

Next, to elucidate the exceptional activity of the in situ generation of dimeric Zn species in PET glycolysis, we performed Hirshfeld and Natural Population Analysis (NPA) charge studies on the rate-determining steps of the intra- and intermolecular bimetallic pathways. − Computational analysis revealed distinct charge redistribution patterns: the intramolecular pathway exhibits attenuated charge fluctuations at the Zn centers compared to the intermolecular system (Table S3). This stabilization of the intramolecular system originates from a bridging hydroxyethoxy anion ligand, which mediates intermetallic charge delocalization and effectively distributes the unfavorable charge build-up occurring during the nucleophilic attack. Critically, this intramolecular dinuclear architecture enables synergistic dual activation via (Figure c): (1) enhanced nucleophilicity of the Zn-bound alkoxide (q _Hirshfeld: −0.200 e vs. −0.187 e; q _NPA: −0.842 e vs. −0.803 e) and (2) increased electrophilicity of the carbonyl carbon (q _Hirshfeld: +0.212 e vs. +0.211 e; q _NPA: +0.893 e vs. +0.871 e). The resultant 11.9 kcal/mol reduction in the rate-determining step barrier (ΔΔG ^‡ = 16.7 kcal/mol to ΔΔG ^‡ = 4.8 kcal/mol) directly correlates with the experimentally observed catalytic enhancement, even under low catalyst concentration conditions (Figure S12).

Based on the aforementioned results, the bimetallic synergistic catalytic mechanism endows the catalytic system with exceptional performance in PET glycolysis, achieving high depolymerization activity while maintaining remarkable BHET selectivity, even at low loadings. This unprecedented efficiency establishes 1 as a prime candidate for the industrial implementation of plastic recycling. Polyesters such as Poly­(trimethylene terephthalate) (PTT), Poly­(butylene terephthalate) (PBT), and Poly­(ethylene naphthalate) (PEN) are derived from various aryl dicarboxylic acids and diols, which expands their range of material applications. However, their compact structures pose significant challenges for depolymerization. Remarkably, 1 achieved a 95% recovery efficiency in the glycolysis of diverse polyesters when employing their corresponding monomer diols, demonstrating its broad substrate compatibility (Figure ). This high efficiency across different substrates relies on a bimetallic synergistic dual-activation strategy, which facilitates the cleavage of strong ester linkages and enables nearly complete recovery of the monomers.

5.

Chemical recycling scope of commercial polyesters via 1-catalyzed glycolysis with a feed ratio of [polymer]₀/[1]₀ =10000/1 at 200 °C.

Polymer blending strategies significantly enhance PET performance and complicate recycling efforts. For example, PET/PE blends improve toughness and crystallization properties, while PET/PP systems enhance viscoelasticity. Interestingly, 1 enabled direct depolymerization of PET in these blends during glycolysis. Moreover, the PE and PP components spontaneously separate during filtration, enabling the simultaneous recovery of PET monomers and PE/PP polymers (Figure a,b and Figure S13). Similarly, PET/PS blends enhance both the barrier and mechanical properties , while still allowing selective glycolysis of PET, thereby enabling the efficient isolation of PS (Figure c and Figure S13). This cascade separation and depolymerization process provides a new approach for the tandem recovery of materials from multicomponent plastics.

6.

Industrial application potential of catalyst 1 for plastic recycling at 200 °C. Catalyst performance in PET glycolysis of mixed polymer wastes (0.02 mol % catalyst loading). (a) PET and PE; (b) PET and PP; (c) PET and PS; (d) Catalyst 1-mediated large-scale PET glycolysis.

To comprehensively evaluate the influence of PET source, morphology, particle size, and mass-transfer effects to strengthen the practical relevance of our catalytic system, we conducted a systematic study encompassing real-world PET waste, various physical forms, particle dimensions, and scale-up reactions. Firstly, four representative post-consumer PET waste sources, bottles, fibers, films, and plates, were collected. Under standardized conditions, all waste types achieved monomer recovery yields exceeding 96% with >99% BHET selectivity (Figure S14), demonstrating the broad substrate tolerance of our catalytic system. Secondly, the effects of PET morphology and particle size were systematically examined using three distinct forms, particles, flakes, and pellets, across varying dimensions (Table S4). Among comparable size scales, particulate PET exhibited the highest activity, followed by flake and then pellet forms (Table S4, Entries 1–6). This trend can be rationalized by differences in specific surface area, which directly govern the mass-transfer efficiency of reactants (e.g., solvent and catalyst) into the polymer bulk. Thirdly, reducing particle size led to a pronounced enhancement in catalytic activity (Entries 1 and 2), a trend consistently observed for both flake and pellet morphologies (Entries 3–6), further underscoring the critical role of mass transport in determining reaction efficiency. Finally, to probe mass-transfer effects under industrially relevant conditions, a scale-up depolymerization was conducted using 20 g of PET, achieving a 98% monomer yield (Figure d). This result confirms that our catalytic system retains high efficiency even at larger reaction volumes. Furthermore, the stability of the catalyst, particularly its susceptibility to deactivation or aggregation during prolonged operation, was investigated in detail. Given the homogeneous nature of our catalytic system, conventional recovery by simple filtration, as commonly employed for heterogeneous catalysts, is not feasible. To simulate practical recycling conditions, we therefore designed a multi-batch continuous glycolysis experiment, in which the catalyst was introduced only at the beginning, followed by sequential addition of fresh PET and EG over five consecutive runs. As shown in Figure , the catalyst exhibited excellent reusability, maintaining BHET yields above 93% across all five batches. These results underscore the high stability and practical viability of our catalytic system for continuous PET glycolysis.

7.

Catalyst stability test for multi-batch continuous glycolysis. Reaction conditions: 0.7 mg of catalyst 1, overall of 0.007 mol % loading, 200 °C, 5 × (1.0 g PET and 5.3 g EG) additions.

Conclusion

In summary, we developed a mononuclear Zn catalyst for efficient PET glycolysis at low loadings (0.01 mol %). Mechanistic analysis suggested that the in situ dimerization of the mononuclear Zn catalyst generates a dinuclear active intermediate, exhibiting exceptional efficiency in PET glycolysis. This process occurs through a bimetallic synergistic catalysis mechanism in which adjacent Zn centers within a single catalytic scaffold cooperatively activate both the nucleophile and the carbonyl carbon via a low-energy barrier pathway. The practical applicability of this system was validated across diverse real-world PET waste sources and physical forms, with systematic studies revealing the critical role of mass transfer in governing catalytic efficiency. A 20 g scale-up depolymerization achieved a 98% monomer yield, and multi-batch continuous glycolysis demonstrated robust catalyst stability with >93% BHET yield over five consecutive runs. Notably, this catalytic system can be extended to the recycling of other polyesters and blended polyester plastics to achieve excellent recycling efficiency. This study demonstrates an efficient and easily implementable protocol for polyester depolymerization, offering a transformative strategy for advancing a sustainable circular plastic economy.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/polymscitech.6c00032.

• Catalytic result of PET glycolysis; mechanistic investigations for PET glycolysis; and NMR spectra of catalyst and depolymerization products (PDF)

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.